Back to Skills

review-research

pjt222
Updated 6 days ago
9 views
17
2
17
View on GitHub
Designaidesign

About

This Claude Skill performs peer reviews of research documents by evaluating methodology, experimental design, and statistical appropriateness. It assesses reproducibility, identifies biases, and provides constructive feedback on manuscripts, preprints, or research proposals. Developers should use it when they need to systematically review the technical quality and rigor of any research-based content.

Quick Install

Claude Code

Recommended
Primary
npx skills add pjt222/agent-almanac -a claude-code
Plugin CommandAlternative
/plugin add https://github.com/pjt222/agent-almanac
Git CloneAlternative
git clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/review-research

Copy and paste this command in Claude Code to install this skill

Documentation

審研

行結構同行審於研工、估法、統擇、可復、總科嚴。

  • 審稿、預稿、內研報
  • 估研提或研協
  • 估稱或薦背證之質
  • 為同事研設於數集前供饋
  • 審論章或論段

  • :研文(稿、報、提、協)
  • :域脈(影法準)
  • :刊或場導(為發審)
  • :補材(數、碼、附)
  • :前審註(若審改版)

一:首過——範與構

讀全文一過、解:

  1. 研問:明且特乎?
  2. 貢稱:何新?
  3. 總構:循期格(IMRaD 或場特)乎?
  4. 範合:工合標群/場乎?
## First Pass Assessment
- **Research question**: [Clear / Vague / Missing]
- **Novelty claim**: [Stated and supported / Overstated / Unclear]
- **Structure**: [Complete / Missing sections: ___]
- **Scope fit**: [Appropriate / Marginal / Not appropriate]
- **Recommendation after first pass**: [Continue review / Major concerns to flag early]

得:明解稿稱與貢。 敗:研問全讀後仍不明→記為主憂續。

二:估法

評研設對域準:

量研

  • 研設應於研問(實、準實、察、查)
  • 樣大有由(力析或實理)
  • 樣法述應(隨、層、便)
  • 變明定(獨、依、控、混)
  • 量具驗、信報
  • 數集程於述可復
  • 倫顧處(IRB/倫准、同意)

質研

  • 法明(紮論、現象、案、人類學)
  • 與選準與飽議
  • 數集法述(訪、察、文)
  • 研者位認
  • 信策報(三角、員察、審跡)
  • 倫顧處

混法

  • 混設由釋
  • 整策述(聚、釋序、探序)
  • 量質二要皆合各準

得:法單畢、各條有特察。 敗:關法訊缺→標主憂、勿設。

三:估統與析擇

  • 統法應於數型與研問
  • 統測設察報(正、同方、獨)
  • 效大與 p 值並報
  • 適處供信區
  • 需處施多較正(Bonferroni、FDR)
  • 缺數處述應
  • 為關設行敏析
  • 果釋合析(不過述發見)

常統紅旗:

  • p-hacking 號(多較、選報、「邊際義」)
  • 不應測(無由於非正數行 t 測、序數行參數測)
  • 混統義與實義
  • 無效大報
  • 後設呈為先設

得:統擇估、特憂書。 敗:審者乏特法專→認、薦專審。

四:估可復

  • 數可性述(開數、庫鏈、請可得)
  • 析碼可性述
  • 軟版環書
  • 隨種或可復機述
  • 關參與超參報
  • 算環述(硬件、OS、依)

可復級:

TierDescriptionEvidence
GoldFully reproducibleOpen data + open code + containerized environment
SilverSubstantially reproducibleData available, analysis described in detail
BronzePotentially reproducibleMethods described but no data/code sharing
OpaqueNot reproducibleInsufficient method detail or proprietary data

得:可復級分含由。 敗:數不可分(私、專)→合成數或詳偽碼為受替——記其供乎。

五:識潛偏

  • 選偏:與代標群乎?
  • 量偏:量程或系扭果乎?
  • 報偏:諸果報乎、含非義者?
  • 確偏:作者唯尋支設證乎?
  • 存偏:脫者、排數、敗實顧乎?
  • 資偏:資源露且可影發見乎?
  • 發偏:此為全像乎、抑負果或缺?

得:潛偏識含稿特例。 敗:自可訊不能估偏→薦作者明處之。

六:書審

建設構審:

## Summary
[2-3 sentences summarizing the paper's contribution and your overall assessment]

## Major Concerns
[Issues that must be addressed before the work can be considered sound]

1. **[Concern title]**: [Specific description with reference to section/page/figure]
   - *Suggestion*: [How the authors might address this]

2. ...

## Minor Concerns
[Issues that improve quality but are not fundamental]

1. **[Concern title]**: [Specific description]
   - *Suggestion*: [Recommended change]

## Questions for the Authors
[Clarifications needed to complete the evaluation]

1. ...

## Positive Observations
[Specific strengths worth acknowledging]

1. ...

## Recommendation
[Accept / Minor revision / Major revision / Reject]
[Brief rationale for the recommendation]

得:審特、建設、引稿確處。 敗:審過長→序主憂、記次患於摘列。

  • 各主憂引特段、圖、或稱
  • 饋建設——患與薦並
  • 正面與憂並認
  • 統估合所用析法
  • 可復明估
  • 薦合所提憂之重
  • 調業、敬、同

  • 糊評:「法弱」無助。明何弱與何故
  • 求他研:審所為研、非汝所為
  • 忽範:會稿期異於刊文
  • 人攻:審工、非作者。永勿引作身
  • 完美:無研全。注於變結之憂

  • review-data-analysis — 深注數質與模驗
  • format-apa-report — APA 格準為研報
  • generate-statistical-tables — 發備統表
  • validate-statistical-output — 統出驗

GitHub Repository

pjt222/agent-almanac
Path: i18n/wenyan-ultra/skills/review-research
0
agentsagentskillsai-assisted-developmentclaude-codeskillsteams

Related Skills

executing-plans

Design

Use the executing-plans skill when you have a complete implementation plan to execute in controlled batches with review checkpoints. It loads and critically reviews the plan, then executes tasks in small batches (default 3 tasks) while reporting progress between each batch for architect review. This ensures systematic implementation with built-in quality control checkpoints.

View skill

requesting-code-review

Design

This skill dispatches a code-reviewer subagent to analyze code changes against requirements before proceeding. It should be used after completing tasks, implementing major features, or before merging to main. The review helps catch issues early by comparing the current implementation with the original plan.

View skill

connect-mcp-server

Design

This skill provides a comprehensive guide for developers to connect MCP servers to Claude Code using HTTP, stdio, or SSE transports. It covers installation, configuration, authentication, and security for integrating external services like GitHub, Notion, and custom APIs. Use it when setting up MCP integrations, configuring external tools, or working with Claude's Model Context Protocol.

View skill

web-cli-teleport

Design

This skill helps developers choose between Claude Code Web and CLI interfaces based on task analysis, then enables seamless session teleportation between these environments. It optimizes workflow by managing session state and context when switching between web, CLI, or mobile. Use it for complex projects requiring different tools at various stages.

View skill